Editing volume and rendering volumes different

Markk655 wrote on 7/20/2010, 7:35 PM
Editing AVCHD 1440x1080 (AC3/5.1SS) in a 1440x1080 project. I use a headset (Logitech G35) to edit to get a feel for the 5.1 SS. However, I can turn that function on/off. I have the problem either way (and via the lousy embedded speakrs in the monitor.

Problem: When I edit, I usually have the music track (typically mp3 source) as background with corresponding volume envelope which relates to video/audio tracks above. When I edit, the music is relatively loud, so I end up bringing the envelope points down to a level that is appropriate for what I would like in the final product. That would be fine, except once I render to AVCHD/.m2ts files (with AC3 audio), the file plays with very low volume for the music soundtrack. Essentially, the volume of the track during editing doesn't match what will be the final output. Is there a way to calibrate this?

One idea I had was to render a few files, load them back into Vegas and see if I can figure out how many db in edit mode = db in rendered product.

Any other (better!) ideas?

Comments

musicvid10 wrote on 7/20/2010, 9:24 PM
I am unable to replicate your observations.

Using the front L-R channels with a stereo WAV reference tone, an AC3 stereo render (AGC off), and a 5.1 AC3 render (AGC off) in VMS 10, the results are all identical, and exactly what I expected (-18dBFS +/- 0.1dB) in each case).

If you used an AGC template, and your aggregate levels were already what you wanted them to be, and they were <0dBFS, then you should use a non-AGC template instead.

As a side note, the phase shift bug I reported with the AC3 Studio preset (when importing a 5.1 AC3 into a 2-channel timeline) is still present, although it was confirmed and a fix was put on the list by Sony almost three years ago. That's a disappointment.
Markk655 wrote on 7/21/2010, 7:56 AM
I agree with you that the output is what is correct. The problem is that I don't know what the output is going to be in terms of soundtrack level based on what I am hearing during editing. When I stick the rendered file back into VMS, it definitely has a lower volume
than what I wanted. The project in edit mode is still louder.

I am not in front of VMS right now, but there are multiple choices for audio preiew drivers. I think the issue maybe choosing the wrong preview audio device driver; or perhaps the headset (which is USB based) does not work well with the Vegas sound as output. I tried different preview device drivers in thye audio setup, but I didn't hear a difference, The headset has minimal configuration possibilities (but I can deactivate the Dolby Surround Sound function and that didn't work)/

So, I'll try a few more things later today. But I was also curious if there is a specific audio selection I should choose (I don't see the headset listed - if I recall therewas a Microsoft Soundmapper, the Realtek HD audio driver from my integrated audio option and I think a Windows option as well). I did a quick check and I didn't see any difference. But maybe I need to a more detailed check. I figure with the USB headset, I shouldn't use my integrated audio driver by default. Is that right? Should I be using or an alternative?

I'll also try another set of headphones (that don't support Dolby and are not USB-based - typical 3.5 mm jack) to see if that helps. I'll also check the AGC (I don't recall that option in VMS, but that could be it!)

Thanks for any and all suggestions.


TOG62 wrote on 7/21/2010, 8:39 AM
From the Help file:

My AC-3 audio files are quieter than my PCM audio files, even though they were rendered from the same project. How can I make the levels match more closely?

The AC-3 encoder can be configured to perform dynamic range compression.

When rendering your AC-3 files, choose the Stereo DVD, AGC on or 5.1 Surround DVD, AGC on template.
musicvid10 wrote on 7/21/2010, 8:54 AM
1) Do your 7.1 headphones show up as a device in Vegas Audio Devices? I know nothing about them or if they are capable of picking up discrete 5.1 from the t/l. A look at the internet says they are a gamer's device that utilizes psychoacoustic modeling. That means various "enhancements" are applied that color the sound, and you are not hearing naked (accurate) 5.1 surround from them.

2) As mentioned, if the levels are correct in the Vegas project (monitoring levels notwithstanding), using AGC is not what one wants to do. AGG compression limits (softens) the louds and raises (loudens) the soft passages. Generally, compression should be applied judiciously at the track level.

3) The OP's problem may also relate to monitoring levels. To be relatively certain of hearing the same levels from the timeline, everything in the chain must be referenced to unity (0dBFS). If anything in the chain is altering the levels, or adding its own compression, the results will not be the same as when rendered. That's a no-brainer.

4) That all being said, playback for monitoring from the timeline under the best of conditions is never quite the same as rendered material. The best one can do is make adjustments in the monitoring chain to compensate, all other things being equal.
Markk655 wrote on 7/21/2010, 11:45 AM
MusicVid, I used to own a headset with true 5.1 (individual speakers in the two sides of the headset) The sound wasn't very good and they broke way too often. So, I went with the Logitech G35, even though it is a USB headset and uses Dolby prologic instead of true/accurate surround sound. However, even when I turn off the Dolby Prologic which should give a standard 2 speaker headset, the level is still not reflective of the rendered file.

I think I have a few things to try The track db slider actually reduces or enhances the volume in the rendered file, right? If nothing else, can I adjust the monitoring level of a specific track within Vegas?
musicvid10 wrote on 7/21/2010, 11:57 AM
Ahh, that explains it. ProLogic surround uses a phase cancelling matrix of a stereo downmix, rather than 6 discrete channels. If you have a card that meets the requirements (see "5.1 Surround Setup" in VMS10 Help), set up six powered speakers in your editing environment. Even with that, I have to do multiple test burns on DVD-RW to get just what I want on the home entertainment system in the living room.

I'm not sure how to manipulate preview volume without affecting output volume in the VMS10 environment. The Preview fader is a stereo downmix, and the lack of Audio Busses precludes routing to anything but single destinations. Maybe there is a trick I haven't discovered yet.
Markk655 wrote on 7/21/2010, 6:57 PM
MusicVid,

I doubt that there is a trick that you haven't uncovered!

Thanks for pointing me in the right direction. It appears to be a VMS 9 to 10 upgrade issue. No, not the software. Just me! I had AGC unchecked in V9 and forgot to check the box in V10. Of course the box was checked .I haven't re-rendered the entire project, but did try out some test tones as the music sound track on a short video 5.1 test. The non-AGC sample was equally loud in edit and after render. As expected, the AGC-checked version was much lower (26 vs 18 db in all 5 speakers (forget LFE).

I never liked re-learning something I already knew. But thankfully we have some very smart people here to remind people like me!

TOG - Thanks for checking into the manual.

As an FYI, as I did try out my alternate headset (std 2 speaker). The volume differences were the same as with the Prologic headset.

Thanks again!